Crucial for social interaction and communication, face recognition undergoes a prolonged maturation process continuing into late adolescence. Existing data suggest that the perception, encoding and retrieval of faces are complex processes, involving activation of widely separated brain regions connected by fibers that undergo increased myelination with age. Thus, we hypothesize that the development of face recognition involves functional & structural maturation of a distributed brain network, including specialized visual cortical areas such as the "fusiform face area" (FFA), as well as medial temporal lobe (MTL) & prefrontal cortex (PFC) regions involved in mnemonic processing. Achieving a new integration of diverse methodologies in pediatric neuro-imaging, we address this hypothesis by applying a combination of behavioral and non-invasive functional & structural methods to subjects ages 7 to 20. Recent event-response potential (ERP) data show specific developmental changes in brain responses that in adults are thought to signify face-category recognition. However, it is not known how these ERP data might relate to category-recognition performance, maturation of FFA, or development of face recognition in children. Thus, we will measure face sensitive ERPs to test correlation with category recognition performance in children versus adults. Using whole-brain fMRI, we will functionally define face selective areas in the occipito-temporal cortex, and test correlations between individual face-recognition performance versus activation in areas involved with face encoding & recognition, namely: the FFA, the MTL and the PFC. For fMRI analysis, we will control for any age dependent changes in the hemodynamic response function (HRF), by measuring each subject's HRF. Given the role of PFC in mnemonic processing and its delayed maturation in children, we will relate behavioral and functional measures of face recognition with DTI measures of frontal white-matter integrity. The proposed combination of ERP and fMRI methods will allow high-resolution analysis of temporal & spatial features of brain activation during face processing. The proposed HRF measures address a key methodological issue in fMRI analysis in children, and DTI will provide a measure of the maturity of the underlying structures. These experiments will break new ground in understanding the development of face recognition in children and will facilitate future research by non-invasive imaging in pediatric populations.